Average velocity tube flowmeter--simple, cheap, energy-saving flow meter

Introduction

Since the advent of the average velocity tube flowmeter in the late 1960s, it has been continuously improved (known abroad as Annubar, Verabar, Probar, Torbar, Itabar, Preso, Deltabar, Averaging Potit-tube, etc.), with different names, but all are based on the principle of pitot tube velocity measurement, which measures the velocity at several points on the diameter (round tube) or length and width (rectangular tube) of the pipeline to calculate the flow rate. Due to its simple structure, easy installation and disassembly, low price and energy saving, it is often used as the preferred instrument in the power, metallurgy, petrochemical and other industries when accurate measurement is not required for trade accounting and it is only used for working condition monitoring, especially when the diameter is greater than 200 mm.

A joint survey of the global flow meter market in the past two years by CONTROL ENGINEERING and Reed Research Group shows that among the 20 commonly used flow meters, the average velocity tube flow meter ranks 8th to 9th; in the century-old project of my country's West-East Gas Transmission, 50 Emerson average velocity tube flow meters were selected on the pipeline with an inner diameter of one meter, accounting for 52% of the total 96 units; in addition, the sales performance of the American Verabar average velocity tube flow meter in the domestic power, metallurgy, petrochemical and other industries is remarkable and impressive. However, domestic average velocity tubes have almost no foothold in the mainland market. The reason is that both Emerson and Verabar have attached great importance to the application of products on site in the past 20 years, spared no effort to innovate and continuously improve; while domestic manufacturers are weak in product development and blindly imitate, and almost produce and sell products that were eliminated more than 10 years ago in foreign countries. These situations are thought-provoking. This article puts forward the following views on several issues that industry insiders have paid more attention to about average velocity tubes in recent years for your reference.

One of the hot topics - the cross-sectional shape of the detection rod

is the hottest topic in the development of the velocity-averaging tube. In the past 30 years, it has been constantly changing and innovating. The more typical ones are as follows:

1. Round

The earliest velocity-averaging tube detection rod had a circular cross section; there were multiple total pressure holes in the upstream direction and low pressure holes in the downstream direction; the middle was separated by a plate. Later, it was believed that since the velocity-averaging tube was in position flow, the static pressure of the entire cross section should be equal, so it was changed to only drill a back pressure hole on the downstream side of the center of the detection rod, and use a thin tube to transmit the back pressure to the low pressure end of the differential pressure transmitter, eliminating the partition and simplifying the structure.

2. Diamond-Ⅰ

In the late 1970s, after years of use, it was discovered that when the Reynolds number was between 105 and 106, the separation point of the fluid on the circular tube would move from 78u to 130u, the so-called "resistance crisis" phenomenon, which changed the pressure distribution on the circular section and caused a flow error of about ±10%. The diamond was gradually replaced. The two sides of the diamond are acute angles, the separation point is determined, and the resistance crisis is eliminated. Other structures remain unchanged.

3. Diamond-Ⅱ

After the diamond-shaped I was used for 7 or 8 years, it was found that the pressure transmission tube of the back pressure hole was easy to be blocked because of its inner diameter of only 3 mm. The American Dieterich Company launched a detection rod section composed of 3 cavities. The total pressure holes were changed from two pairs to 3 to 4 pairs. The back pressure holes corresponded to the total pressure holes one by one, and the total back pressure lead-out pipe was cancelled. This structure will not affect the normal operation of the average velocity tube even if one or two back pressure holes are blocked. [page]

4. Wing

In the past 20 years, people have been introducing various detection rod shapes with low resistance from the perspective of reducing the resistance of the averaging tube, such as symmetrical airfoils, oblate circles, ellipses (Preso), etc. In fact, the permanent pressure loss of the averaging tube is only a few dozen Pa, which can be ignored and there is no need to make a fuss. However, this type of cross-sectional shape has the disadvantage of low output differential pressure, which is not worth the cost. However, there are special cases. Emerson uses this wing-shaped profile structure to measure steam. Due to the high steam flow rate and high density, it is possible to obtain a large differential pressure. It is indeed necessary to reduce the resistance to increase the strength, but it is limited to one model and is used in special occasions.

5. Bullet shape

The American Verabar Company has introduced a rough surface treatment on the front end of the bullet (roughness X/KS is about 200), claiming that this can control the thickness of the boundary layer, thereby improving the measurement accuracy. In actual estimation, the influence of the boundary layer on the accuracy is negligible. The low pressure of the bullet type is taken from both sides, and the output differential pressure is 20%~30% smaller than that of the diamond, round and T-shaped, which is not conducive to the selection of low gas flow rate.

6. T-shaped

This is a new structure introduced by Emerson in the past two years (the company calls it 485 Annubar). There are two rows of total pressure holes on the T-shaped detection rod facing the flow direction, and two rows of back pressure holes in the vortex area facing the back flow direction. Emerson claims that due to its innovative slot design, the accuracy will be improved; and the back pressure is in the T-shaped vortex area, which can increase the output differential pressure by about 20% compared with the diamond and circular shapes. Multiple low-pressure holes are used on the back. The total pressure and back pressure holes of this structure are less than 2 mm, which is easy to clog and can only be used for clean fluids.

7. Delta

Deltaflow averaging tube was launched by Systec Co., Germany. It claimed to have many advantages at the MICONEX 2004 exhibition, but there is no essential difference from the diamond-II in terms of cross-sectional shape and structure. It is still an insertion flow meter, and it cannot get rid of the basic mode of measuring the flow velocity at a few points on the diameter of the pipeline to infer the flow rate. The manufacturer advertises that its straight pipe section is only 3~7D, and the accuracy can reach ±0.6%, which is unconvincing and unbelievable. But its material selection is worth looking at. Generally, averaging tube materials are mostly 316 stainless steel; while Deltaflow uses 1.4528 or Hastelloy alloy steel, which can withstand temperatures as low as -200℃ and as high as more than 1000℃, and can be applied to various corrosive media.

Hot spot 2 - the number and position of detection holes

The averaging tube is an insertion flow meter with sampling properties. In the early days, more than ten points of total pressure were measured in the diameter direction of the averaging tube, but no matter how many measuring points there were, it could only reflect the flow velocity distribution on a certain diameter, not the entire cross section.

The premise for sampling to be meaningful is that there is a straight pipe section of 20~30D (D is the inner diameter of the pipe) in front of the velocity-averaging pipe. At this time, the flow in the pipe is fully developed, and the velocity lines of the velocity distribution are concentric circles symmetrical to the axis, that is, the velocity at the same radial direction is equal. Only in this way, measuring only a few points of velocity on the diameter can reflect the situation of the entire cross section. There are many different opinions on which points of velocity on the diameter to measure, and there are roughly the following 5 kinds (see Table 1).

The velocity distribution in Table 1 is based on the fully developed turbulent mathematical model proposed by Nikuradse. In the early 1990s, the study of pipe flow showed that although the Nikuradse formula was simple, the fully developed turbulent flow described by it was quite different from the actual situation at both the pipe wall and the center of the pipe (especially near the pipe wall), so the total pressure holes should be increased to 3 pairs, and 4 pairs are currently used for larger pipe diameters. Its distribution is based on the logarithmic-Chebyshev method (see Table 2), and has been confirmed by ISO TC30:

As for the position and number of low-pressure holes, since the velocity-averaging tube is in a potential flow, there is no lateral flow on the cross section, and the static pressure at each point is equal, so one point or multiple points can be selected. There is no problem of accuracy. It is just that multiple points are not easy to be blocked, and the pressure difference obtained from the back of the detection rod will be greater than that obtained from both sides. [page]

For a long time, people have believed that the total pressure measured by each total pressure hole reflects the flow velocity distribution in the pipeline. Due to the unequal flow velocity, the measured total pressure is also unequal. The pressure output after these total pressures are "averaged" in the high-pressure cavity of the velocity-averaging tube is the total pressure of the average flow velocity in the pipeline. In fact, this is not the case. Due to the unequal total pressure at each point, there will be flow in the high-pressure cavity, and even eddy currents at the edge of the hole, which will cause pressure loss. In 1975, William.H. et al. conducted tests and research on this and proposed an empirical formula, but there are still some coefficients in this empirical formula that need to be determined by experiment, and the average flow velocity differential pressure value cannot be directly calculated. Therefore, to date, no matter what methods are used to determine the position and number of total pressure holes, the average velocity tube still needs to be calibrated experimentally to determine the flow coefficient K. As the flow rate changes (i.e., the change of Re), the change of the velocity distribution near the pipe wall will be greater than that of the center of the pipe. The logarithmic-Chebyshev total pressure distribution has one more measuring point near the pipe wall to adapt to this change, which is more reasonable.

Hot spots that are easy to ignore - pipelines

For more than 30 years, manufacturers have spared no effort on the above two issues and made a lot of innovations. But as far as the average velocity tube itself is concerned, it is only a multi-point flow meter. It can only be called a flow meter if it is inserted into the pipeline to measure the flow. Moreover, the influence of the pipeline on the flow measurement of the average velocity tube is very important and cannot be ignored. It is mainly manifested in the following two aspects:

1. The length of the straight pipe section

The length of the straight pipe before the average velocity tube must reach 20~30D to ensure that the velocity distribution is fully developed turbulent. Only in this way can the flow through the entire cross section be estimated by measuring the velocity at only a few points. Otherwise, the flow in the pipeline will be as shown in Figure 3, which is relatively complicated (the situation is similar after other resistance parts). The velocity distribution is not only asymmetrical to the axis, but also has lateral flow (secondary flow) and vortex. What can be explained by measuring the velocity at a few points on the diameter? How to ensure the accuracy of the measurement?

2.

The average velocity tube of the inner diameter of the pipeline can only measure the velocity. To measure the flow rate, the pipe cross section must be multiplied (the inner diameter of the circular pipe must be measured, and the width and height of the rectangular pipe must be measured). It is also an insertion instrument. In practical applications, it is often difficult or not to measure the inner diameter seriously.

ISO7145 believes that when the inner diameter cannot be measured, it is allowed to use a soft ruler to measure the circumference of the outer diameter of the pipeline and estimate the wall thickness to determine the inner diameter. Of course, this cannot confirm the influence of corrosion and fouling on the inner wall of the pipeline. How can the inner diameter of the pipe determined in this way be guaranteed to be accurate? From the following analysis, it can be seen that the accuracy of the inner diameter of the pipe will become a crucial factor affecting the accuracy of the average velocity tube flow measurement. If

the estimation of flow accuracy

only considers the main factors, the average velocity tube calculation formula can be simplified to:

①In the formula, QV is the volume flow rate; C depends on the coefficient of each parameter unit; D is the inner diameter of the pipe; DP is the output differential pressure; x is the fluid density. The flow uncertainty derived from formula ① is

From the above, the manufacturer has made unremitting efforts on the shape and measuring point position of the average velocity tube detection rod. They will only affect the size of the output differential pressure Dp. From formula ②, it can be seen that the relative error sD/D of the inner diameter D of the tube will have an impact on the flow accuracy that is several times greater than the error sDp/Dp of the differential pressure Dp.

In addition, even if the flow coefficient K in formula ① is calibrated by the manufacturer one by one, it is obtained under specific conditions in the laboratory, and the flow field conditions of the laboratory are often not met on site. At this time, the use of the flow coefficient K provided by the manufacturer will also bring about a large error. As early as 20 years ago, W. Rahmeyer and C. L. Britton installed average velocity tubes at 2 to 12D after the resistance parts (elbows, gate valves, etc.) and conducted systematic experimental tests. The test shows that when the straight pipe section is less than 4 to 5D, the deviation of the flow coefficient will reach more than ±8%. The

above analysis further illustrates that the pipeline (including the inner diameter D and the length of the straight pipe section L) is the main factor affecting the accuracy of the average velocity tube flow measurement! The length of the straight pipe section

in the industrial control system

depends on the process requirements. It is difficult to arrange a straight pipe section of more than 20D to accommodate the flow meter. Without a long enough straight pipe section, the average velocity tube must face the harsh fact that the error may reach ±8%! Is there still a place for the average velocity tube in the industrial control system?

The average velocity tube is only a detection link that provides information sources in the industrial control system. The flow information reflected by its output should accurately reflect the change of flow, not necessarily the exact value. Here, the importance of repeatability is often greater than accuracy. For example, in the combustion regulation control system of the boiler, the air flow should be measured, and the size of the fuel should be adjusted and controlled to ensure the best combustion efficiency. Here, as long as the output of the average velocity tube has a definite single-valued function relationship with the flow, and does not change arbitrarily, that is, the repeatability is good, it is fine. As for the exact value of the air volume (accuracy), it is not a problem that the system is concerned about. In other words, as long as it does not involve logistics settlement (trade, economic evaluation, etc.), accuracy is not a major consideration.

Research on the average velocity tube over the past 30 years has shown that when the straight pipe section does not meet the requirements, its error can reach more than ±8%, but the repeatability can often reach 0.5%. As long as the purpose of use is not trade settlement (such as natural gas metering), but for the regulation and monitoring of industrial control systems, the average velocity tube has its advantages of simple structure, especially in large-diameter situations, and is often used as the preferred instrument.

Problems faced by development and innovation

The average velocity tube has been in existence for more than 30 years and has occupied a place in the flow family with its advantages such as simple structure. However, as the saying goes, "blessings are accompanied by misfortunes; misfortunes are accompanied by blessings", these advantages inevitably bring it the following three disadvantages:

1. Low accuracy

Over the years, nearly 20 types of average velocity tubes have been developed. Because they are inserted, they can only reflect the flow velocity through the detection rod. No matter how many measuring points are taken on it, they can only reflect the flow velocity distribution on the diameter (or width, height) of the pipe section. When the straight pipe does not meet the requirements, these points lose their representative meaning and the accuracy is difficult to be better than ±3%.

2. Small output differential pressure

The average velocity tube is based on the principle of pitot tube velocity measurement. It measures the total static pressure to calculate the flow rate. It is often used for large-caliber gas measurement. At this time, the output differential pressure is only tens of Pa (several millimeters of water column). This is determined by its principle and structure. Although the manufacturer has spared no effort to make a fuss on the detection rod for many years, the harvest is limited. The latest T-shaped structure, even if the manufacturer says that the output differential pressure is increased by 20%, the increased output differential pressure is of no help from a practical point of view.

3. Easy to block

Since the flow must be measured through the detection hole, as long as there is dust, solid particles, condensate, etc. in the fluid, blockage is difficult to avoid. Although it is easy to disassemble and even can be repaired without flow, it is not a good thing and is difficult for users to accept.

The developers of the average velocity tube have made unremitting efforts for more than 30 years to address the above problems and have made many improvements, but if they do not jump out of the old model, it will be difficult to make breakthrough progress, and it seems that there is no way out. But if you open your mind and learn from other instruments to make up for your shortcomings, can you see a new spring? The author has made an attempt to do this and obtained a patent in 1986 (CN852045298). It has some effect, but it is not obvious. Recently, I have made some improvements on this basis and am applying for a patent, hoping to do my best to promote the application of the average velocity tube. (end)